The QuickFeather AVS Reference Design integrates the Alexa voice-initiated Close-Talk experience, enabling a broad set of battery-powered applications to communicate directly with Alexa for a multitude of use cases. It is offered as the first available, Close-Talk qualified smart hearable reference design based on the company's EOS S3 ultra-low power Arm Cortex-M4-based Voice Processor + eFPGA System-on-Chip.
The reference design is built on the company's Open Reconfigurable Computing (QORC) initiative, and its QuickFeather 100% open source development kit. The EOS S3 integrates Sensory's Low Power Sound Detection (LPSD) technology, with DSP Concepts' TalkTo noise suppression and beamforming technology, and Alexa Wake Word engine technology to enable a superior user experience and long battery life.
"The age of the smart hearable has arrived," says Brian Faith, CEO of QuickLogic, "and our always-listening, Alexa Close-Talk reference design is the ideal platform for developing next generation AI-enabled hearable products. OEMs and ODMs can use this readily available design along with the QuickFeather development kit as a sandbox for their own product development with the comfort that we have already gone through the AVS certification process on our reference design with Amazon."
"The fact that we have built all of this on our QuickLogic Open Reconfigurable Computing platform, which integrates with the SensiML AI Software Platform," says Faith, "makes it that much easier for users to innovate in familiar development environments, speeding time-to-market and reducing risk."
In addition to the very low power, voice-initiated Alexa voice recognition capability enabled by the EOS S3 platform, OEMs and ODMs can optionally integrate multiple AI-based motion, sound and other AI use cases based on SensiML's AI Software Platform, which already has native support for the QuickFeather development kit.
The QuickFeather AVS Reference Design Features:
- Alexa Wake Word Engine (WWE) running in the EOS S3 voice processor
- Hardware-optimized Low Power Sound Detection (LPSD) technology from Sensory
- Support for one or two microphone use cases with DSP Concepts' TalkTo